Application of genetic algorithms to combinatorial synthesis: A computational approach to lead identification and lead optimization

Journal of the American Chemical Society

Volumn 118, Issue 7, 1996, Pages 1669-1676

  Singh, Jasbir ^a,h   Ator, Mark A ^a,f   Jaeger, Edward P ^a,c   Allen, Martin P ^a,d   Whipple, David A ^a,d   Soloweij, James E ^a,e   Chowdhary, Swapan ^a,g   Treasurywala, Adi M ^a,b  

^a Sanofi Winthrop ^*  (United States)

^b NPS Pharmaceutical Inc   (United States)

^c JOHNSON AND JOHNSON   (United States)

^d PFIZER INC   (United States)

^e AMGEN INC   (United States)

^f CEPHALON INC   (United States)

^g Sanofi Winthrop Inc   (United States)

^h Nycomed Inc   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

STROMELYSIN;

ALGORITHM; ARTICLE; DRUG ANALYSIS; DRUG DEVELOPMENT; ENZYME ASSAY; ENZYME SUBSTRATE; FLUORESCENCE; MODEL; PEPTIDE SYNTHESIS; QUANTITATIVE STRUCTURE ACTIVITY RELATION;

EID: 0029878670     PISSN: 00027863     EISSN: None     Source Type: Journal    
DOI: 10.1021/ja953172i     Document Type: Article

References (54)

1. For an excellent review see: (a) Gallop, M A , Barrett, R W.; Dower, W. J.; Fodor, S P. A., Gordon, E. M. J Med. Chem. 1994, 37, 1233-1251 (b) Gordon, E. M.; Barrett, R. W Dower. W J.; Fodor, S. P A., Gallop, M. A. J. Med. Chem. 1994, 37, 1385-1401 and references cited therein.
2. For an excellent review see: (a) Gallop, M A , Barrett, R W.; Dower, W. J.; Fodor, S P. A., Gordon, E. M. J Med. Chem. 1994, 37, 1233-1251 (b) Gordon, E. M.; Barrett, R. W Dower. W J.; Fodor, S. P A., Gallop, M. A. J. Med. Chem. 1994, 37, 1385-1401 and references cited therein.
3. Pirrung, M. C. Chemtracts. Org. Chem 1994, 7, 184-186
4. Martin, E. J.; Blaney, J. M.; Siani, M. A : Spellmeyer, D. C., Wong, A. K.; Moose, W H J. Med Chem. 1995, 38, 1431-1436.
5. See ref 3 for indices used to represent dissimilarity.
6. There is tremendous effort in producing and screening chemically diverse compound libraries. However, the real interest for the pharmaceutical industry is the biological diversity that is embodied in it. i e. biodiversity not necessarily chemodiversity.
7. This approach would not require explicit description of a set of indices, but instead, the selection of indices would be implicit in this type of optimization approach.
8. For general references for genetic algorithms see: (a) Holland, J H. Sci Am. 1992 66 (b) Forrest. S. Science 1993, 261, 872-878
9. For general references for genetic algorithms see: (a) Holland, J H. Sci Am. 1992 66 (b) Forrest. S. Science 1993, 261, 872-878
10. (a) Wagener, M.; Gasteiger, J. Angew, Chem., Int. Ed. Engl. 1994, 33, 1189-92.
11. (b) Walters, D. E., Hinds, R. M. J. Med. Chem. 1994, 37, 2527-2536.
12. (c) Wehrens, R.; Lucasius, C.; Buyden, L.; Kateman, G. Anal. Chim Acta 1993, 277, 313-324.
13. (d) See ref 9 in ref 6a (listed above) for use of genetic algorithms for jet engine design.
14. (a) Judson, R. S., Jaeger, E. P., Treasurywala, A. M.; Peterson, M. L J. Comput. Chem. 1993, 14, 1407-1414.
15. (b) Judson, R. S.; Jaeger, E. P . Treasurywala, A. M. J. Mol. Struct (THEOCHEM) 1994, 308, 191-206.
16. See refs 10-16 in ref 8a above.
17. Singh, J.; Allen, M. A.; Ator, M. A.; Gainor, J. A.; Whipple, D. A . Soloweij, J. E . Treasurywala, A. M : Morgan, B. A.; Gordon, T. D.; Upson, D. A. J Med Chem. 1995, 38, 217-219.
18. We use all 20 coded amino acids (see a complete list under Abbrevations). except Cys. We employ S-methylcysteine (Smc. denoted by the single letter code U1 as the 20th amino acid
19. We do not use deletion and insertion as we do not want to change the overall size of (he chromosome and. therefore, overall length of the bit strings.
20. Crossover is the single most important aspect which provides for most optimum assurance to explore the gene population for selecting a set of more fit members.
21. In a true sense this random generation should be referred to as generation 0 (zero) as far as GA's are concerned, since there are no fitness functions which need 10 be evaluated by GA to provide the initial population of members.
22. Genesis version 1.2 from ftp site: ftp atc nrl navy mil.
23. 5 (i.e., 32) bits. Some of the 20 amino acids are represented by more then one string The choice of this bit degeneracy was selected at random, but once selected, it was kept constant through out the experiment.
24. Where X represents one ammo acid present at a time and the numbering is used for the sake of discussion only.
25. (a) Ator, M.; Beigel, S.; Dankanich, T.; Echols, M : Gainor, J.; Gilliam, C.; Gordon, T.; Koch, D.; Kruse, L : Morgan, B.; Olsen, R.; Siahaan, T.; Singh, J.; Whipple, D. Peptides: Chemistry Structure and Biology: Proceedings of the 13th American Peptide Symposium: Hodges, R., Smith, J., Eds.; 1994: pp 1012-1016
26. (b) We have previously described the reasons and the validation for use of CPG as a solid support, see ref 10 above for details.
27. (a) We have incorporated an alanine residue at the N-terminus of all sequences identified by GA before we tag the N-terminus with the fluorescent marker - COP. This was carried out to distance the marker group further away from the active site of a protease.
28. (b) Gainor, J. A.; Gordon, T. D.; Morgan, B. A. Peptides: Chemistry Structure and Biology. Proceedings of the 13th American Peptide Symposium: Hodses, R., Smith, J., Eds.; 1994. pp pp 989-991.
29. The synthesis outlined here does not involve any "mix & split" strategy, and therefore, each sample/tube represents a single sequence
30. The weighing robot was programmed such that any sample which falls outside this weight range was reweighed. Each rack of assay also contained a negative control. hexa-D-alanyl, and GPLAMF as a positive control sample. These samples are used as a guide to decide the validity of every assay and are used as an integral part of biological evaluation.
31. (a) Chowdhury, S. K., Vavra, K. J.; Brake, P. G.; Banks, T.; Falvo, J.; Wahl, R., Eshraghi, J.; Gonyea, G.; Chait, B. T.; Vestal, C H Rapid Commun. Mass Spectrum. 1995, 9(7), 563-569
32. (b) Brownell, J.; Earley, W.; Kunec, E.; Morgan, B. A., Olyslager, B.; Wahl, R. C., Houck, D. R. Arch. Biochem. Biophys. 1994, 314, 120-125.
33. (a) Eshraghi, J.; Chowdhury, S. K. Anal. Chem 1993, 65, 3528-3533.
34. (b) Following is a typical sample procedure for obtaining data from LC/MS: The procedure adopted for on-line separation using a microcapillary HPLC system coupled to an electrospray lonization (ESI) mass spectrometer has been described in details in the above reference. Briefly, the gradient mobile phases (0.1% aqueous trifluoroacetic acid (TFA) and 0 1% TFA in acetonitrile) from the Waters 600 HPLC pump (200 μL/min) is split with a ratio of 100:1. The smaller fraction (2 μL/min) passes through a 0.5 μL injection loop followed by a microcapillary column (VYDAC C-18, 300 Å; 300 mm id ×15 cm) to first to the microdetection cell of a Spectroflow UV detector and then to the electrospray lonization chamber of a Finnigan TSQ 700 mass spectrometer (Finnigan Mat, San Jose, CA). while the larger fraction goes to the waste The UV detector and the ESI mass spectrometer are operated in series so that measurement of the UV chromatogram and the mass spectra can be performed on the same effluents. A sheath liquid (2-methoxyethanol) was added to the LC effluents prior to electrospray lonization at a flow rate of 2 μL/min to assist the electrospray lonization
35. The observed (raw) fluorescence values were used as a fitness function. The normalized data have been used here to compare various generations The data were normalized such that the activity of each positive control sample represents 10000 fluorescence units.
36. The genetic algorithm we have used here is very similar to the one utilized previously for computational experiments related to conformational analysis problems, where the GA's goal was to identify conformations with lower energy (i.e., the fitness function was used to compute energy): therefore, for this experiment, we hase used the converse, i.e negative. value as a fitness function to optimize activity.
37. However, in a given generation (except of course the initial "random" generation), there are invariably some sequences which are already present in earlier generation(s). We have decided to re-test these samples for the number of occurrences asked for by GA. These multiple results take into account the variance in the biological assays. In addition, these results also provide an indication of the robustness and rehability of our assay.
38. pop in my generation)
39. For example, a 60-70% probability for occurrence of a given group should provide a better method to introduce constrains. This is analogous to the approach used for phage display based experimental design for introducing constrains at a given position(s) by using nucleotides NNN. where N represents 70:10:10:10 proportions of different nucleotide monomers. For a representative reference on phage display approach to introduce constrains. see: Schatz, P. J. Biotechnology 1993, 11, 1138-143.
40. 1 site
41. For sake of clarity only selected samples are shown. A complete list of sequences and their normalized data is available as supporting information.
42. Since we have determined the site of processing only for a handful of samples in any given GA-based generations, we have not used site of cleavages as an input for obtaining the Gen : set in this study
43. Various classes of MMP inhibitors have been reported in the literature, (a) For hydroxamate series, see: (i) Singh, J.; Conzentino, P.; Cundy, K.; Gainor, J.; Gordon, T.; Johnson, J.; Morgan, B.; Whipple, D.; Gilliam, C.; Schneider, E.; Wahl, R. BioMed. Chem. Lett. 1995, 5, 537-542. (ii) Johnson, W. H.; Roberts, N. A.; Borkakoti, N. J. Enzyme Inhib. 1987, 2, 1-22. (b) For 5-carboxyalkyl series, see: Chapman, K. T.; Kopka, I. E.; Durette, P. L ; Esser, C. K.; Lanza, T. J.; Izquierdo-Martin, M.; Neidzwiecki, L.; Change, B.; Harrison, R. K.; Kuo, D. W.; Lin, T.; Stein, R. L. J. Med. Chem. 1993, 36, 4293-4301. (c) For phosphonate series, see: (i) Bartlett, P. A., Marlowe, C. K. Biochemistry 1987, 26, 8553. (ii) Bird, J.; DeMallo, R. C.; Harper, G. P.; Hunter, D. J.; Karran, E. H.; Maekwell, R. E.; Miles-William, A. J., Rahman, S S.; Ward, R. W. J. Med. Chem. 1994, 37, 158-169.
44. Various classes of MMP inhibitors have been reported in the literature, (a) For hydroxamate series, see: (i) Singh, J.; Conzentino, P.; Cundy, K.; Gainor, J.; Gordon, T.; Johnson, J.; Morgan, B.; Whipple, D.; Gilliam, C.; Schneider, E.; Wahl, R. BioMed. Chem. Lett. 1995, 5, 537-542. (ii) Johnson, W. H.; Roberts, N. A.; Borkakoti, N. J. Enzyme Inhib. 1987, 2, 1-22. (b) For 5-carboxyalkyl series, see: Chapman, K. T.; Kopka, I. E.; Durette, P. L ; Esser, C. K.; Lanza, T. J.; Izquierdo-Martin, M.; Neidzwiecki, L.; Change, B.; Harrison, R. K.; Kuo, D. W.; Lin, T.; Stein, R. L. J. Med. Chem. 1993, 36, 4293-4301. (c) For phosphonate series, see: (i) Bartlett, P. A., Marlowe, C. K. Biochemistry 1987, 26, 8553. (ii) Bird, J.; DeMallo, R. C.; Harper, G. P.; Hunter, D. J.; Karran, E. H.; Maekwell, R. E.; Miles-William, A. J., Rahman, S S.; Ward, R. W. J. Med. Chem. 1994, 37, 158-169.
45. Various classes of MMP inhibitors have been reported in the literature, (a) For hydroxamate series, see: (i) Singh, J.; Conzentino, P.; Cundy, K.; Gainor, J.; Gordon, T.; Johnson, J.; Morgan, B.; Whipple, D.; Gilliam, C.; Schneider, E.; Wahl, R. BioMed. Chem. Lett. 1995, 5, 537-542. (ii) Johnson, W. H.; Roberts, N. A.; Borkakoti, N. J. Enzyme Inhib. 1987, 2, 1-22. (b) For 5-carboxyalkyl series, see: Chapman, K. T.; Kopka, I. E.; Durette, P. L ; Esser, C. K.; Lanza, T. J.; Izquierdo-Martin, M.; Neidzwiecki, L.; Change, B.; Harrison, R. K.; Kuo, D. W.; Lin, T.; Stein, R. L. J. Med. Chem. 1993, 36, 4293-4301. (c) For phosphonate series, see: (i) Bartlett, P. A., Marlowe, C. K. Biochemistry 1987, 26, 8553. (ii) Bird, J.; DeMallo, R. C.; Harper, G. P.; Hunter, D. J.; Karran, E. H.; Maekwell, R. E.; Miles-William, A. J., Rahman, S S.; Ward, R. W. J. Med. Chem. 1994, 37, 158-169.
46. Various classes of MMP inhibitors have been reported in the literature, (a) For hydroxamate series, see: (i) Singh, J.; Conzentino, P.; Cundy, K.; Gainor, J.; Gordon, T.; Johnson, J.; Morgan, B.; Whipple, D.; Gilliam, C.; Schneider, E.; Wahl, R. BioMed. Chem. Lett. 1995, 5, 537-542. (ii) Johnson, W. H.; Roberts, N. A.; Borkakoti, N. J. Enzyme Inhib. 1987, 2, 1-22. (b) For 5-carboxyalkyl series, see: Chapman, K. T.; Kopka, I. E.; Durette, P. L ; Esser, C. K.; Lanza, T. J.; Izquierdo-Martin, M.; Neidzwiecki, L.; Change, B.; Harrison, R. K.; Kuo, D. W.; Lin, T.; Stein, R. L. J. Med. Chem. 1993, 36, 4293-4301. (c) For phosphonate series, see: (i) Bartlett, P. A., Marlowe, C. K. Biochemistry 1987, 26, 8553. (ii) Bird, J.; DeMallo, R. C.; Harper, G. P.; Hunter, D. J.; Karran, E. H.; Maekwell, R. E.; Miles-William, A. J., Rahman, S S.; Ward, R. W. J. Med. Chem. 1994, 37, 158-169.
47. Various classes of MMP inhibitors have been reported in the literature, (a) For hydroxamate series, see: (i) Singh, J.; Conzentino, P.; Cundy, K.; Gainor, J.; Gordon, T.; Johnson, J.; Morgan, B.; Whipple, D.; Gilliam, C.; Schneider, E.; Wahl, R. BioMed. Chem. Lett. 1995, 5, 537-542. (ii) Johnson, W. H.; Roberts, N. A.; Borkakoti, N. J. Enzyme Inhib. 1987, 2, 1-22. (b) For 5-carboxyalkyl series, see: Chapman, K. T.; Kopka, I. E.; Durette, P. L ; Esser, C. K.; Lanza, T. J.; Izquierdo-Martin, M.; Neidzwiecki, L.; Change, B.; Harrison, R. K.; Kuo, D. W.; Lin, T.; Stein, R. L. J. Med. Chem. 1993, 36, 4293-4301. (c) For phosphonate series, see: (i) Bartlett, P. A., Marlowe, C. K. Biochemistry 1987, 26, 8553. (ii) Bird, J.; DeMallo, R. C.; Harper, G. P.; Hunter, D. J.; Karran, E. H.; Maekwell, R. E.; Miles-William, A. J., Rahman, S S.; Ward, R. W. J. Med. Chem. 1994, 37, 158-169.
48. The sequence: GPST-YT. chosen for P1′ variations was observed in all five generations. This sequence happens to be processed selectively by stromelysin.
49. - ions.
50. Singh, J., Ghose. A. Unpublished results.
51. Translation of the stromelysin selective P1′ and P2′ information to a series of inhibitors using solid phase based combinatorial synthesis is obviously the next logical step It is feasible to involve GA to facilitate in rapid resolution to the selective inhibitor identification/optimization problem.
52. Computationally it has been shown that, in general. GA should converge in about 10- 15 generations (see ref 8a). This would mean that one would have to synthesize a total of ∼600 samples, i.e. <0.002% of the combinatorial population of 3 200 000. One could easily synthesize these relatively small numbers of compounds as individual compounds (via parallel synthesis) and would not have to resort to preparing mixtures. This in turn would obviate considerations of alternate decoding strategies to identify active compounds.
53. q, represent pendants on a core scaffold or a hybrid of these. Here, n, m, p, and q represent number of variables at the respective positions. These templates may represent peptidomimetic, peptoid, or small molecule based templates for lead selection/optimization either as protease inhibitors or for receptor antagonists.
54. The templates recently described by Martin et al. (ref 3 above) and Kick and Ellman (Kick, E. K.; Ellman, J. A. J. Med. Chem. 1995, 38, 1427-1430) are the two relevant examples to which the current version of the genetic algorithms may be potentially applied for the identification/ optimization of lead compounds, respectively.